Thread Reader helps you read and share Twitter threads easily!
I'm @ThreadReaderApp a Twitter bot here to help you read threads more easily. To trigger me, you just have to reply to (or quote) any tweet of the thread you want to unroll and mention me with the "unroll" keyword and I'll send you a link back on Twitter 😀
— Thread Reader App (@threadreaderapp) November 25, 2017
X thread is series of posts by the same author connected with a line!
From any post in the thread, mention us with a keyword "unroll"
@threadreaderapp unroll
Follow @ThreadReaderApp to mention us easily!
Practice here first or read more on our help page!
Recent
Fuel Explosion Under Bridge of Americas in Panama
The following is a summary of information provided by AI (@Grok):
The 3 fuel tanker trucks that exploded yesterday below the Bridge of the Americas in Panama, have a max combined capacity of about 30,000 gals.
The following is a summary of information provided by AI (@Grok):
The 3 fuel tanker trucks that exploded yesterday below the Bridge of the Americas in Panama, have a max combined capacity of about 30,000 gals.
For security reasons, the combined max capacity of the FIXED gas/diesel storage tanks @ this site (POTSA) isn't publicly known, but best estimates put that number @ around 75 MILLION gallons.
The storage facility existed prior 2 the construction of the bridge & no evidence of prior safety concerns related 2 this storage facility being located below the bridge. The likelihood of a chain reaction (domino effect) from 1 fixed storage tank exploding @ this terminal is low
Eating 1000 Sardines Gave Me THIS Superpower
(New 2026 Findings!)
1/8) I ran a self-experiment where I ate 1000 sardines in a month.
Sure, it made me stink—but it also gave me one epic superpower. Let me explain. 🧵 (link at the end)
We all know sardines make your breath stink and that they’re nutrient-dense.
That’s basic.
But eating that many sardines changed me. It gave me a “superpower” that had my inner Marvel nerd activated—and my scientist brain scrambling to explain it.
Eventually, I found those data.
(New 2026 Findings!)
1/8) I ran a self-experiment where I ate 1000 sardines in a month.
Sure, it made me stink—but it also gave me one epic superpower. Let me explain. 🧵 (link at the end)
We all know sardines make your breath stink and that they’re nutrient-dense.
That’s basic.
But eating that many sardines changed me. It gave me a “superpower” that had my inner Marvel nerd activated—and my scientist brain scrambling to explain it.
Eventually, I found those data.
2/8) It was new paper in a top journal turned confusion into clarity and left me in awe of how much we’re still uncovering about human physiology.
The superpower…
Full deep dive link: staycuriousmetabolism.substack.com/p/why-stinking…
The superpower…
Full deep dive link: staycuriousmetabolism.substack.com/p/why-stinking…
3/8) I became cold resistant. The effect wasn’t subtle. I could stand shirtless in a Boston blizzard without so much as a goosebump.
9,000 robotic missions in one month — Ukraine is automating the frontline🧵
In March alone, Ukraine’s ground robotic systems (UGVs) executed 9,000+ missions — up from ~2,900 in Nov 2025 and 7,000 in Jan 2026.
Since the start of 2026: 32,500 missions — @DefenceU .
This is not experimentation anymore. It’s scaling.
Why it matters?👇
In March alone, Ukraine’s ground robotic systems (UGVs) executed 9,000+ missions — up from ~2,900 in Nov 2025 and 7,000 in Jan 2026.
Since the start of 2026: 32,500 missions — @DefenceU .
This is not experimentation anymore. It’s scaling.
Why it matters?👇
2 / What’s happening
Ukraine is systematically shifting high-risk battlefield tasks to robots:
— ammunition resupply
— logistics under fire
— casualty evacuation
Already 167 units are using UGVs (vs 67 in late 2025).
All missions are tracked, verified, and scored via the DELTA battlefield system — turning robotics into a measurable, managed capability.
Ukraine is systematically shifting high-risk battlefield tasks to robots:
— ammunition resupply
— logistics under fire
— casualty evacuation
Already 167 units are using UGVs (vs 67 in late 2025).
All missions are tracked, verified, and scored via the DELTA battlefield system — turning robotics into a measurable, managed capability.
3 / What it means (operational level)
1. Casualty reduction as a system goal
Robots are replacing soldiers in the most dangerous segments of logistics. This is a structural shift — not tactical improvisation.
2. Industrialised battlefield learning
Through DELTA, every mission becomes data.
→ feedback loops
→ rapid iteration
→ performance benchmarking across units
3. Logistics becomes semi-autonomous
The frontline is evolving from “human supply chains” to robot-assisted sustainment networks.
1. Casualty reduction as a system goal
Robots are replacing soldiers in the most dangerous segments of logistics. This is a structural shift — not tactical improvisation.
2. Industrialised battlefield learning
Through DELTA, every mission becomes data.
→ feedback loops
→ rapid iteration
→ performance benchmarking across units
3. Logistics becomes semi-autonomous
The frontline is evolving from “human supply chains” to robot-assisted sustainment networks.
1/8 Nod (@nizzaneela) has updated his paper demolishing Pekar et al's claim of 'two introductions'.
v3 is more accessible, same brutal conclusion: the supposed support for two spillovers is an artefact of rigged hypothesis testing.
arxiv.org/html/2502.2007…
v3 is more accessible, same brutal conclusion: the supposed support for two spillovers is an artefact of rigged hypothesis testing.
arxiv.org/html/2502.2007…
2/8 Quick recap: Pekar et al claimed a Bayes Factor of 60 for two introductions.
After three major errors were found (including a bad cut-and-paste!), it dropped to 4, just at the noise level border.
After three major errors were found (including a bad cut-and-paste!), it dropped to 4, just at the noise level border.
3/8 Nod later showed that the 4 is still wrong.
The one-introduction model was tested against stricter conditions than the two-introduction model, which is basic statistical malpractice.
Under the same conditions, the Bayes factor is <0.25.
The one-introduction model was tested against stricter conditions than the two-introduction model, which is basic statistical malpractice.
Under the same conditions, the Bayes factor is <0.25.
To those who are heroes let’s work — we are death dealers and if we die we will be resurrected again!
Death to all spies! Death to those who commit treason to those in war! No, no, no. I do not wish death on Any one.
I wish long life on you Elon Musk @elonmusk so stop
Death to all spies! Death to those who commit treason to those in war! No, no, no. I do not wish death on Any one.
I wish long life on you Elon Musk @elonmusk so stop
Interfering in our elections.
Your child’s situation is in God’s hands. Stop chasing money man! Turn to God! Jesus forgives!
Serve God Elon! Repent, and to recompense and atonement for God. Start with the Bible. Start with the two great commandments and see where that leads you
Your child’s situation is in God’s hands. Stop chasing money man! Turn to God! Jesus forgives!
Serve God Elon! Repent, and to recompense and atonement for God. Start with the Bible. Start with the two great commandments and see where that leads you
My phone is always open —- 678-233-7061
You still owe me money —- and if you refuse to pay me how many others have you stolen from
If you are okay with stealing from #HiramHeroes you are okay with stealing votes too?
You still owe me money —- and if you refuse to pay me how many others have you stolen from
If you are okay with stealing from #HiramHeroes you are okay with stealing votes too?
World remember Elon says “incentives drive outcomes”
Now we know why @DolezalForGA was not stopping your votes from getting flipped!
Who will stop the votes from being flipped #GaPol
Any new candidate wants to be a hero ?
@BerniceKing please view 👆
Now we know why @DolezalForGA was not stopping your votes from getting flipped!
Who will stop the votes from being flipped #GaPol
Any new candidate wants to be a hero ?
@BerniceKing please view 👆
They have been cheating and we have submitted the proof! So now let’s pray for arrest! They have been doing Republican on Republican and democrat on democrat cheating too!
Collecting lots of money in poly markets!
So now we know how Elon got so rich!
Collecting lots of money in poly markets!
So now we know how Elon got so rich!
He has been betting on elections!
Then they hack the system and declare the winner so they can maximize their odds!
@FBI @NAACP @JasmineForUS @KamalaHarris did you all know? @ikamalhaasan my brother from PSR, did you know this breaking news? Elon’s apolymarket.com/event/ga-13-de…
Then they hack the system and declare the winner so they can maximize their odds!
@FBI @NAACP @JasmineForUS @KamalaHarris did you all know? @ikamalhaasan my brother from PSR, did you know this breaking news? Elon’s apolymarket.com/event/ga-13-de…
6 BEHAVIOURS THAT INSTANTLY KILL ATTRACTION :
(stop doing these 🚫)
Must read thread 🧵
(stop doing these 🚫)
Must read thread 🧵
I GAVE CLAUDE MY SALARY AND IT GAVE ME A PLAN.
Use these 15 prompts to turn your paycheck into a strategy:
Use these 15 prompts to turn your paycheck into a strategy:
1 | Salary Allocation Architect
"Act as a personal finance strategist. My monthly income is [SALARY], my fixed expenses are [EXPENSES], my current savings are [SAVINGS], and my financial goal is [GOAL]. Create a smart allocation plan for spending, saving, and investing."
"Act as a personal finance strategist. My monthly income is [SALARY], my fixed expenses are [EXPENSES], my current savings are [SAVINGS], and my financial goal is [GOAL]. Create a smart allocation plan for spending, saving, and investing."
2 | Money Flow Analyzer
"My monthly income is [SALARY] and my expenses are [EXPENSE LIST]. Evaluate how my money moves each month and recommend a more efficient structure that improves my savings rate."
"My monthly income is [SALARY] and my expenses are [EXPENSE LIST]. Evaluate how my money moves each month and recommend a more efficient structure that improves my savings rate."
7 Types of Women No Man Should EVER Marry
(She’ll paralyze your mind,
Drain your confidence,
And make you forget you were ever a man)
(She’ll paralyze your mind,
Drain your confidence,
And make you forget you were ever a man)
Most men don’t die from bullets in war anymore.
They die in bedrooms...
Next to women they never should’ve trusted.
She looked soft.
Spoke sweet.
Touched him like forever.
Now?
He wakes up to stress.
Walks on eggshells in his own home.
Sleeps beside a woman who hasn’t touched him in months.
Not because he’s weak…
But because he chose wrong.
He didn’t marry a wife.
He married a trap with mascara.
And every day?
He dies a little more inside.
Not every woman is wife material.
The wrong one won’t just ruin your life… She’ll reprogram your soul.
Here are the 7 types of women who destroy a man from the inside out 👇🏽
They die in bedrooms...
Next to women they never should’ve trusted.
She looked soft.
Spoke sweet.
Touched him like forever.
Now?
He wakes up to stress.
Walks on eggshells in his own home.
Sleeps beside a woman who hasn’t touched him in months.
Not because he’s weak…
But because he chose wrong.
He didn’t marry a wife.
He married a trap with mascara.
And every day?
He dies a little more inside.
Not every woman is wife material.
The wrong one won’t just ruin your life… She’ll reprogram your soul.
Here are the 7 types of women who destroy a man from the inside out 👇🏽
1. The Sex Weaponizer
(She doesn’t use sex to connect. She uses it to train you like a dog.)
In the beginning, she gives it freely.
Wild.
Relentless.
It feels like love.
But it isn’t.
It’s leverage.
Once she senses he’s hooked?
She slows down.
Pulls back.
Turns intimacy into a negotiation.
He starts folding.
Apologizing.
Tiptoeing around her moods for a breadcrumb of affection.
And from that moment on…
He’s no longer a man in love.
He’s a man in training.
Many men don’t realize it.
But they’re not being loved…
They’re being conditioned.
(She doesn’t use sex to connect. She uses it to train you like a dog.)
In the beginning, she gives it freely.
Wild.
Relentless.
It feels like love.
But it isn’t.
It’s leverage.
Once she senses he’s hooked?
She slows down.
Pulls back.
Turns intimacy into a negotiation.
He starts folding.
Apologizing.
Tiptoeing around her moods for a breadcrumb of affection.
And from that moment on…
He’s no longer a man in love.
He’s a man in training.
Many men don’t realize it.
But they’re not being loved…
They’re being conditioned.
🧵 MemPalace claims to be "the highest-scoring AI memory system ever benchmarked" I cloned it. Installed it. Ran the benchmarks. Read every line of code. Here's what's actually inside. A thread.
1/ The headline claim: 96.6% on LongMemEval, beating Mem0 (~85%), Zep (~85%), Mastra (94.87%)
Problem: MemPalace reports retrieval Recall@5. The others report end-to-end QA accuracy.
Those numbers measure completely different things. You can't put them in the same table.
Problem: MemPalace reports retrieval Recall@5. The others report end-to-end QA accuracy.
Those numbers measure completely different things. You can't put them in the same table.
2/ LongMemEval_s has ~50 sessions per question.
MemPalace retrieves at session granularity with n_results=50 (all of them). So R@5 asks: "is the right session in your top 5 out of ~50?"
Random baseline: 10%
Any decent embedding model: 95%+
This is a trivially easy retrieval setting.
MemPalace retrieves at session granularity with n_results=50 (all of them). So R@5 asks: "is the right session in your top 5 out of ~50?"
Random baseline: 10%
Any decent embedding model: 95%+
This is a trivially easy retrieval setting.
I've seen a fair few people suggest recently that the 'science of reading' can essentially be boiled down to two messages:
1. Explicitly teach kids to decode (i.e. systematic phonics).
2. Build knowledge.
But in many schools, these two things alone will fail.
>>
1. Explicitly teach kids to decode (i.e. systematic phonics).
2. Build knowledge.
But in many schools, these two things alone will fail.
>>
In England, explicit phonics and the sequencing of knowledge have been prioritised for a decade.
Despite valuable national progress, I can say confidently that it's perfectly possible for a school to have these things in place and yet fail to secure strong reading outcomes.
>>
Despite valuable national progress, I can say confidently that it's perfectly possible for a school to have these things in place and yet fail to secure strong reading outcomes.
>>
This isn't me naysaying the value of teaching phonics explicitly or building pupils' knowledge of the world.
On the contrary, both of these things are essential, and I spend plenty of my professional life explaining to teachers and school leaders exactly why I think this.
>>
On the contrary, both of these things are essential, and I spend plenty of my professional life explaining to teachers and school leaders exactly why I think this.
>>
